4 Methods of analysis

4.1 Morphological features

K' images of the galaxies in grey scale with isocontours overlaid are shown in Figs. 1a-28a. All figures are only available in electronic form.

In order to detect the presence of features showing a departure from radial symmetry such as bars and spiral arms, we have used a masking technique. We filtered the original images with a box of 2-3 times the FWHM of the seeing (median filter) and divided the observed image by the filtered one. The resulting images (hereafter called the sharp-divided images to differenciate them from those obtained by subtraction, i.e. the so-called unsharp masking technique) are shown in Figs. 1b-28b. This technique is very well suited to trace asymmetries in the light distribution, such as bars, spiral arms, dust lanes, rings; it allows the subtraction of the diffuse background in a very convenient way to look for subtle, small-scale variations and discuss the possible presence of both dust extinguished and more luminous regions (Sofue et al. 1994; Márquez & Moles 1996; Márquez et al. 1996; Erwin & Sparke 1999; Laine et al. 1999).

The presence of bars can be determined quantitatively by studying the behaviour of isophotal position angles (PAs) and ellipticities ($\epsilon$s): a bar is characterized by a local maximum in the $\epsilon$ corresponding to a constant PA (Wozniak et al. 1995; Friedli et al. 1996; Jungwiert et al. 1997). Therefore, we have fit ellipses to the isophotes (Jedrzejewski 1987) with the IRAF ellipse task in stsdas.analysis.isophote, allowing the center, PA and $\epsilon$ to vary from one isophote to the next. Foreground stars have been masked out. The resulting parameters allow the reconstruction of a model galaxy. The difference between the original image and the model (hereafter called the difference images) are particularly useful to trace all the features that cannot be described with ellipses, as spiral arms and boxy or peanut-like components. They are given in Figs. 1c-28c and discussed separately for each galaxy. When the PA suddenly changes from one isophote to the next, the model cannot work for the regions left inside (see Figs. 1c, 8c, 9c, 11c, 12c, 15c, 18c, 19c, 25c and 26c).

4.2 Surface photometry and photometric decomposition

The major axis PA (counted anticlockwise from north to east as usual) and $\epsilon$ from the ellipse fitting are drawn as a function of the ellipse semi-major axis (therefore, in radius) in Figs. 1e-28e.

In Figs. 1f-28f the surface brightness profiles (isophotal magnitude fit to the ellipse versus ellipse semi-major axis) are given for K' and J. We have used the 1-D resulting profiles to obtain the bulge and disk contributions. We have fit an exponential to the outermost region (the disk), subtracted it to the observed profile and fit a r^1/4 law (the bulge) to the residual; this calculation was continued until convergence was achieved (see Márquez & Moles 1996, 1999). The resulting bulge and disk parameters are given in Table 3. The surface brightness profiles in J and K' are plotted in Figs. 1f-28f with the bulge, disk and bulge+disk fits superimposed. The corresponding residuals are shown in Figs. 1g-28g. Since residual background subtraction is the biggest source of error in determining the profiles (see for instance de Jong 1996), we caution the reader that in the cases where the galaxy occupies most of the frame and the residual background level could only be determined in very small regions, the error in the profiles can reach 15-20%. Otherwise, residuals are in general smaller than 20%, except in the bar and spiral arm regions.

   

Table 3: Bulge+disk decomposition parameters and magnitudes

Galaxy	$\mu_{{\rm d}}$	$r_{{\rm d}}$	$m_{{\rm d}}$	$\mu_{{\rm b}}$	$r_{{\rm b}}$	$m_{{\rm b}}$	$m_{\rm T}$	$\mu_{{\rm d}}$	$r_{{\rm d}}$	$m_{{\rm d}}$	$\mu_{{\rm b}}$	$r_{{\rm b}}$	$m_{{\rm b}}$	$m_{\rm T}$	$m_{ J}^{{\rm ap}}$	$r_{{ J}}^{{\rm ap}}$	$m_{K'}^{{\rm ap}}$	$r_{K'}^{{\rm ap}}$
	J							K'
(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)	(14)	(15)	(16)	(17)	(18)	(19)
UGC 1395	-	-	-	-	-	-	-	19.8	21	10.5	16.3	1.17	12.6	10.4	-	-	10.82	35
IC 184	-	-	-	-	-	-	-	18.2	14	9.9	17.7	4.9	10.9	9.5	-	-	9.98	35
IC 1816	18.8	11	11.0	11.2	0.2	10.9	10.2	18.1	13	9.8	12.6	0.54	10.56	9.38	10.65	45	9.52	45
UGC 3223	-	-	-	-	-	-	-	18.1	20	8.9	15.9	2.1	10.9	8.7	-	-	9.28	45
NGC 2639	17.6	17	8.7	18.2	9.6	9.9	8.4	16.7	18	7.8	17.5	14	8.3	7.3	9.08	50	7.83	50
IC 2510	19	19.2	10.1	18.0	2.9	12.3	10.	17.9	16	9.1	10.6	0.3	10.2	8.8	10.82	42	9.56	42
NGC 3281	19.6	92	7.03	18.8	15	9.48	6.9	19.1	109	6.2	17.8	17	8.3	6.1	9.42	35	8.43	35
NGC 3660	21.2	44	10.3	21.4	57	9.3	8.9	19.1	18	10	19.8	19	10	9.4	10.63	35	9.89	20
NGC 4253	18.4	10	10.7	17.3	2.7	11.7	10.3	17.0	10	9.4	6.4	0.1	7.3	7.2	10.63	50	9.21	50
NGC 4507	20.1	31	9.9	17.9	7.5	10.1	9.3	18.9	32	8.7	16.0	4.8	9.2	8.2	9.97	44	8.75	44
NGC 4785	17.1	22	7.7	12.3	0.7	9.7	7.6	17.5	25	7.8	15.0	2.1	9.9	7.6	9.73	28	8.77	28
NGC 5347	19.8	23	10.3	19.1	7.9	11.2	9.9	18.2	19	9.1	15.5	2.2	10.4	8.8	10.30	48	9.00	48
NGC 5728	20.2	103	7.4	17.0	6.1	9.6	7.3	18.1	22	8.8	15.3	3.6	9.1	8.2	9.58	38	8.68	38
ESO 139	20.0	25	10.3	15.6	1.3	11.6	10.	19.1	26	9.3	15.2	1.7	10.6	9.0	10.54	45	9.69	45
NGC 6814	19.4	62	7.7	18.5	9.2	10.3	7.6	18.0	40	7.3	16.6	5.5	9.5	7.1	9.30	47	8.04	47
NGC 6860	18.2	11	10.3	13.2	0.4	11.5	10.	18.1	15	9.6	16.6	3.9	10.3	9.1	10.56	38	9.49	38
NGC 6890	18.4	16	9.7	17.8	3.5	11.7	9.6	17.2	15	8.6	15.5	1.8	10.8	8.5	10.35	29	8.97	38
NGC 6951	20.4	69	8.5	19.4	22	9.3	8.1	18.6	55	7.2	17.4	16	8.0	6.8	8.6	112	7.20	112
NGC 151	23.8	191	9.7	19.6	25	9.3	8.7	19.3	11	11.4	19.2	15	10.0	9.7	9.18	42	8.68	45
IC 454	19.4	53	8.1	18.2	5.8	11.0	8.0	18.1	74	6.1	16.6	5.5	9.5	6.0	10.02	45	8.77	45
NGC 2712	19.0	24	9.5	10.2	0.3	9.4	8.7	18.3	26	8.5	13.5	1.5	9.3	8.1	9.69	55	8.59	55
NGC 2811	19.3	47	8.2	17.4	9.8	9.0	7.8	17.4	26	7.6	14.6	3.6	8.4	7.2	9.07	38	7.92	38
NGC 3571	20.1	68	8.3	18.4	9.0	10.3	8.1	19.3	50	8.1	17.3	7.9	9.4	7.8	10.09	38	9.13	38
NGC 3835	18.8	25	9.1	16.9	2.7	11.4	8.9	17.8	23	8.3	18.4	12	9.7	8.0	10.22	55	8.86	55
NGC 4162	18.7	17	9.8	17.3	2.7	11.8	9.6	17.5	18	8.6	13.7	0.9	10.5	8.4	9.78	55	8.64	55
NGC 4290	18.6	20	9.4	13.7	0.7	11.2	9.2	17.6	20	8.3	12.6	0.7	9.9	8.1	9.74	60	8.46	60
NGC 4779	20.4	37	9.8	16.0	1.8	11.4	9.6	19.4	37	8.9	14.8	1.7	10.2	8.6	10.06	65	8.83	65
NGC 6012	19.1	30	9.1	16.6	0.8	13.8	9.0	18.0	30	7.9	11.8	0.2	11.6	7.8	10.00	48	8.90	48
NGC 6155	19.2	19	10.2	20.3	11.8	11.5	9.9	18.5	18	9.5	20.7	52	9.2	8.6	10.29	55	9.19	55

(1) Galaxy name.
(2) Equivalent surface brightness for the disk component in J band.
(3) Equivalent radius for the disk component in J band.
(4) Computed total magnitude for the disk in J band.
(5) Equivalent surface brightness for the bulge component in J band.
(6) Equivalent radius for the bulge component in J band.
(7) Computed total magnitude for the bulge in J band.
(8) Computed total bulge + disk J magnitude.
(9)-(15) The same as (2)-(8) for the K' band.
(16) J magnitude measured from simulated aperture photometry.
(17) Radius of the circular aperture used to measure the J magnitude.
(18) K' magnitude measured from simulated aperture photometry.
(19) Radius of the circular aperture used to measure the K' magnitude.
Surface brightnesses are given in magnitudes arcsec^-2 and radii in arcsec.

4.3 Color images and color gradients

For galaxies for which a J band image is also available, we display in Figs. 1d-28d the ratio of the J to K' images (after background subtraction), median filtered by 3$\times$3 pixels in order to enhance the S/N of the outer regions. Color images will allow a more complete description of the previously detected morphological features in terms of the contribution of dust and/or star forming events. We note that in the cases where the background is not flat, somewhat structured color images are obtained (galaxies with non-flat remaining backgrounds are indicated with an asterisk in Col. 1 of Table 2).

Magnitudes integrated in circular apertures and J-K' color gradients were obtained from the curves of growth as in Márquez & Moles (1996). The apertures and the corresponding magnitudes are given in Table 3. Color gradients are shown in Figs. 1h-28h for galaxies for which we have data in the two bands. The background subtraction was done as described above and the same caveat applies, in the sense that some color gradient could be artificially produced for the galaxies with non-flat backgrounds.

   

Table 4: Total magnitudes for the 9 galaxies with also K' and/or Jmagnitudes from the litterature

Galaxy	$K'_{{\rm us}}$	$K'_{{\rm others}}{\rm (ref)}$	$J_{{\rm us}}$	$J_{{\rm others}}{\rm (ref)}$
UGC 1395	10.82	10.3(2)
NGC 2639	7.83	8.44(5)	9.08	9.40(5)
NGC 3281	8.43	9.12(1)	9.42	10.28(1)
NGC 4253	9.21	9.9(2),9.56(5),9.65(6)	10.63	11.04(5),11.01(6)
NGC 4507	8.75	9.28(1),8.9(4)	9.97	10.27(1)
NGC 5347	9.00	9.7(2),9.8(4)
NGC 5728	8.68	9.21(1)	9.58	10.25(1)
NGC 6814	8.04	8.65(1),8.0(4)	9.30	9.72(1)
NGC 6890	8.97	9.26(1),8.0(4)	10.35	10.31(1)
NGC 6012	9.02	9.19(3)

(1) K magnitudes from Glass & Moorwood (1985) (2) McLeod & Rieke (1995) (3) de Jong (1996) (4) Alonso-Herrero et al. (1998) (5) Hunt et al. (1999) (6) K' magnitudes from Mulchaey et al. (1997).

   

Table 5: Parameters of the bar(s)

Galaxy	Primary	Primary	Primary	Detection*	Secondary	Secondary	Secondary	Detection*
	bar PA	bar	bar size		bar PA	bar	bar size
		$\epsilon$	(arcsec)			$\epsilon$	(arcsec)
UGC 1395	145	0.56	16	f,o	139	0.32	7	s,h
IC 184	170	0.50	9	i,f,d	30	0.40	4	f,h
IC 1816	110	0.50	10	i,s,f,d	0:	0.20:	1:	c,h
UGC 3223	75	0.71	21	i,s,f,d	75:	0.45:	5:	f,h
NGC 2639	137	0.40	8	f,s	117	0.15	2	s,h
IC 2510	145	0.61	15	i,s,f	143:	0.52:	7:	s
NGC 3281	133	0.40	5	i,s,o
NGC 3660	110	0.70	23	i,s,f,d
NGC 4253	105	0.50	11	i,s,f,d	5	0.10	2	s,d,h
NGC 4507	52	0.35	9	s,d,f,o
NGC 4785	65	0.50	10	i,f	82	0.35	5	f,d,c,h
NGC 5347	105	0.60	34	i,d,f
NGC 5728	30	0.60	44	i,f,o	90	0.45	4	i,s,f,d
ESO 139-12
NGC 6814	25	0.30	12	i,s,f,o
NGC 6860	10	0.55	10	i,s,f	90	0.10	4	f,d,s,c,h
NGC 6890	15	0.30	6	s,f,o
NGC 6951	84	0.52	45	i,s,f,d,o
NGC 151	155	0.45	18	i,s,f,d	80	0.30	5	s,f,d,h,c
IC 454	128	0.76	18	i,s,f,d	133	0.40	3	s,f,c
NGC 2712	25	0.60	20	i,d,f	5	0.25	5	s,f,d,c
NGC 2811	30	0.50	20	s,f	20	0.30	5	s,f,d
NGC 3571	90:	0.25:	2:	s,c,h,f
NGC 3835	55	0.40	5	s,f,c
NGC 4162	145:	0.30:	12:	f,s
NGC 4290	35	0.65	26	i,s,f,d	55:	0.35:	6:	s,f,d
NGC 4779	8	0.65	30	f,d
NGC 6012	155	0.65	16	i,s,f,d
NGC 6155	160:	0.35:	15:	i,f	120	0.40	6	s,f,d

^* We refer to the image in which the bar is detected as following: i: direct image; s: sharp-dividing; f: ellipse fitting parameters; d: difference image (original - model); c: color image; h: HST image; o: others (see text).

Notes:
All the elongated central structures we call secondary bars could be either bars or elongated rings or disks; kinematical data will help to discriminate among them.
The quantities followed by a column are those for which we are not certain that such an elongation is detected.

In Table 4 we give the magnitudes that we measure for 9 galaxies, together with J, K' and/or K magnitudes found in the literature. A straightforward comparison between our J values and those reported in previous works (for the same apertures) gives: $J_{{\rm us}} - J_{{\rm other}} = -0.42~\pm~0.25$. Only 4 galaxies in our sample have been observed in K' by Mulchaey et al. (1997); we find: $K'_{{\rm us}} - K'_{\rm Mulchaey} = 0.015 \pm 0.63$. If we exclude NGC 6890 (see Sect. 5.17) we obtain: $K'_{{\rm us}} - K'_{\rm Mulchaey} = -0.30~\pm~0.36$. For 10 galaxies and a total of 12 measurements, we find: $K'_{{\rm us}} - K_{{\rm other}} = -0.42~\pm~0.32$. This value is comparable to the difference between $K'_{\rm Mulchaey}$ and $K_{{\rm other}}$ ( $-0.55~\pm~0.49$). These differences in the zero points are not critical for our purposes and will not be discussed further. In any case, we notice that the differences in the colour indexes J-K' are within the errors.

In Table 5 we give the parameters we determine for primary and secondary bars in the sample galaxies. We specify which of the methods above have been used to detect the presence of secondary bars (or primary bars when detected for the first time), and if evidence is found in available HST images. We note that, excepting the secondary bar of IC 2510, only seen in the sharp-divided image (just marginally in the PA-$\epsilon$plot), the secondary bars are obtained in at least two methods.